Also, does their PHEVx range rating explain how Ford is using the term “equivalent electric range” for their being tested JCI-Saft PHEV?

As to emissions: They are at least aware of the challenges and are considering testing separately in each mode. I think that you are right – a PHEV with a prolonged AER (like the EREV Volt) would do better on overall emissions than a PHEV with a blended cycle.

Nonetheless, my observation so far is that converting EV mode and hybrid mode into mpge ratings just isn’t done (yet).

For example, in evaluating performance of PHEVs, there’s this document:

http://www.transportation.anl.gov/pdfs/HV/399.pdf

They analyzed plug-in conversions, and they looked at electricity consumption and gasoline consumption as totally separate quantities. Almost like a two-part efficiency rating. Google.org does the same thing for their plug-in initiative.

However, there’s more to this than mpg. Emissions control is also important.

The emissions of the plug-in Priuses actually went UP versus the production version. This was because the engine was doing lots of starts/stops and under more infrequent conditions, due to the increased use of the electric motor. As a result, the catalytic converter could never warm up properly, and it couldn’t control the emissions as normal.

They did look at a quasi-series PHEV from Renault – the only test subject that wasn’t based a parallel hybrid architecture. I say quasi because the ICE couldn’t sustain the battery at a certain SOC -it just made it drain more slowly. But I got the sense that they ignored emissions on the Renault because it was a Euro city car platform that would never come to the US anyway.

I think looking at emissions of a series hybrid plug-in vs parallel plug-in would be very interesting. If the series’ engine is subjected to fewer start/stops, and is on for a more sustained duration, its emissions might be superior to the parallel.

Just my two cents. let me know what you think.

The papers themselves are pdfs. Maybe someone like AES can explain it better to me, but it seems that the methods are still a work in progress. This is what I can figure out from it anyway. They are very aware of the issue of variable mpg based on mileage traveled:

“The caution from EPA that ‘your mileage may vary’ takes on a whole new meaning if your driving style and especially your daily driving distances can change your fuel economy by a factor of 2-3. The changes are remarkably distorted when a range-extender PHEV is considered. Consider the specifications of the GM Volt concept PHEV; your mileage may vary – from 50 MPG (driving long distances, or you forgot to charge) to infinity (no fuel consumed makes an undefined MPG level).”

Their method seems to be to use a “mileage weighted probability curve” which they also refer to as a “utility factor (UF)” to determine for fleet wide use what percent of travel is actually in charge depleting mode (all EV in the case of the Volt) vs in charge sustaining mode, and weight those together.

So if I understand this right (and please help me AES or anyone else better at this my now aching brain is!) then the Volt’s EPA MPG for CAFE purposes would be something like this according to their example curve: the percent of the fleet’s time in EV only times the zero plus the percent spent in charge-depleting time the gallons used per mile which is 1/50 or 0.02:
(0.60*0g/m + .40*0.02g/m)/1 = 0.008g/m = 125mpg rating. (That’s if I read that curve right as meaning that they believe that 60% of daily driving would be under 40 miles range and 40% over that. I wonder if I got that wrong though as I thought a much greater percent was under 40 miles/day … again help from those with better minds for this than mine would be appreciated.)

I noticed the same thing.

Lyle:

What’s that? Just curious. I checked back and it seemed that my comments went up all right, so no problem, but it did seem strange.

“Measuring MPG is going to become a really murky concept – how will miles per gallon equivalent (MPGE) be factored into the equation?”

MPG may be hard to figure out for a particular vehicle, especially over a short interval, for the reasons given. However, for GM or some other company the bedrock concept is clear enough: MPG is total miles traveled (of the cars sold) divided by total gallons used (by the cars sold). Determining a numerical value requires some estimates but is no more elusive than what is done now, and maybe less so.